June 1999
Towards an Enterprise View of Information in Clinical Chemistry
Arden W. Forrey. PhD
University of Washington Seattle WA

This newsletter report to the LISMI Division membership is intended to convey a perspective about the direction of information management capabilities in the clinical laboratory within healthcare. Over the past several years, the evolving complexity of the recognized role of information within healthcare has lead to the need to consider an "Enterprise View" in the way that the meaning is conveyed by the information technology is documented, transformed, into requirements for architectures and then implemented using available technologies. Enterprises need to be defined in terms of purposes, stakeholders, obligations, responsibilities among stakeholders, and the processes, data and constraints involving the information needed to achieve those purposes and meet the obligations and responsibilities. If the parts of the enterprise are defined and documented separately, such as in the clinical laboratory, this must be done so that the separate parts can be "composed" into a larger whole in a consistent fashion. In the past his has not been done and the consequences of not even attempting to do so are becoming unavoidable. Full interoperability is not yet possible. Technical interoperability of platforms is coming close but content interoperability is still out of reach because the concepts composing the underlying meaning are inconsistent and incomplete. The present chaotic state of healthcare vocabularies is one example of the problem in data representation. The status of data structure compatibility is equally confused. There is need for notations for representing a logically consistent definition of an enterprise which deals with not only the syntax but also the underlying semantics that reflects the meaning in a way that can validate the logical consistency of the definition. This is termed a "Domain Information Model"". We are at the stage now of recognizing the need to do this in order to get at the "Clinical Decision Support" capabilities for care functions and "Management Decision Support" function for Resource Management and Analysis functions if the work using these information services is to be verifiably valid.

So, how is the LISMI Division proceeding to deal with this need? First, by working within the voluntary consensus efforts in the US to help evolve US national standards that deal with the situation in the US, the stage is now set to also be able to participate, via the international standards activities, in developing a consistent position with the general conventions needed for the global market. The recent American Medical Informatics Association Fall Symposium (8-10 November 1998) Workshop on the G8 activities in healthcare information clearly reveals the extent of cross-border healthcare services and the attendant information component. The US-Canada example is illustrative. Enterprise conventions will therefore need to have a common international core. Since 1997 the AACC LISMI Division has endeavored to develop close working relationships with the Association of Clinical Biochemists in the UK via their Information Technology Group. There are already numerous important points that have been learned from this interaction that will help assemble an Enterprise View and use it in the definition of clinical lab services. In ASTM E-31.13 an evolving draft standard with this scope is already in an online form at the ASTM Web site. Individuals who have an interest in commenting or contributing should contact the author by e-mail at forraw@u.washington.edu . This document will be a resource for the work being done by Dr. Jay Jones in E-31.13 on defining the approach to data management for POCT in follow up to the workshop held by the POCT Division at the 1998 AACC Annual meeting. He can be reached at JBuJones@PSGHS.edu .It is important to know that NCCLS is also joining the ANSI HISB as a member beginning in 1999. Thus the work that the Area Committee on Laboratory Automation is doing will fit into the evolving mosaic of standards that will lead to an Enterprise View and Domain Information Model applying to the clinical laboratory. These, and other steps leading to Clinical Views of the Electronic Health Record (EHR), will bring to the Division membership a base of knowledge and tools which can be used is in local environments to build the Enterprise View. For example, The Standard Guide to the Diabetes View of the EHR being worked on by Dr. Harry Wetzler who can be reached at HWetz@aol.com . These are but a few of the activities that will contribute to an Enterprise View that can be the basis for further extension. At a future time the LISMI Division will have a workshop on approaches to implementing such interoperable Enterprise Views.

Page Access: